Settlements in Private Antitrust Litigation
Files
Description
In this chapter a simple theory of settlement behavior is used to analyze the data from the Georgetown project on private antitrust litigation. The empirical version of the model is used to simulate the effect on settlements of a reduction in the damage multiplier from its current level of three. The simulations indicate that reducing the damages multiplier in antitrust cases could dramatically lower the settlement rate as well as affect the propensity of plaintiffs to bring suits and to win suits that go to trial. In recent years there has been substantial attention paid by lawyers, politicians, and academics to the volume and form of private antitrust litigation. Treble damages has been a focal point because there has been serious concern as to whether the benefits of discouraging inappropriate business behavior outweigh the costs of the litigation process. Public discussion about the costs of litigation often focuses on the cost of trial even though a substantial portion of cases end without a court trial. Most of the academic literature on antitrust deterrence and litigation has also generally ignored settlement as well—at least as far as questions of optimal damages and choice of damage multiplier are concerned. Rather, the literature has focused primarily on the deterrence associated with the payment of compensatory damages by a defendant when a case is lost. By building on recent theoretical work on the economics of the litigation process and on data from the Georgetown project, this chapter expands current knowledge about antitrust settlements. Section 4.2 provides a nontechnical description of a framework for describing and evaluating the antitrust litigation process. Litigation consists of three interdependent strands: first, the decision by the plaintiff whether to file suit; second, the decisions by plaintiff and defendant whether to settle; and third, the decisions by both parties as to what type of effort to put forth during the trial. The process is inherently interdependent because the decision to sue depends on one’s expectations as to the likelihood of settlement and trial and on the gains or losses associated with each. Likewise, expectations concerning the effort to be put forth at trial affect the probability of litigation and settlement. Thus, while the second strand of the litigation process (settlement) is emphasized, it is essential that the analysis take into account the relationship between settlements and the decision to bring suit and to expend effort at trial. In appendix A the three strands of the litigation process are examined in more detail. In section 4.3 the empirical analysis is presented. Settlement rates are shown to vary systematically with the nature of the case and the stakes involved. For some policy purposes it is important to look at conditional settlement rates – how settlement rates vary when one variable of interest changes while other variables are held fixed. Therefore we examine conditional settlement rates, using a statistical model to evaluate the probability that cases settle. We also simulate the effects of a change in the damage multiplier on the settlement rate. In appendix B the process of simulation is described in greater detail. Section 4.4 contains some brief concluding remarks.
Source Publication
Private Antitrust Litigation: New Evidence, New Learning
Source Editors/Authors
Lawrence J. White
Publication Date
1988
Recommended Citation
Perloff, Jeffrey M. and Rubinfeld, Daniel L., "Settlements in Private Antitrust Litigation" (1988). Faculty Chapters. 1867.
https://gretchen.law.nyu.edu/fac-chapt/1867
